Impact prediction, evaluation and mitigation - Introduction to Environmental Impact Assessment - page 139

The various impact identification methods discussed in Chapter 4 may also be of value in
impact prediction. The Sassaman threshold-of-concern checklist has already been noted;
the Leopold matrix also includes magnitude predictions, although the objectivity of a
system where each analyst is allowed to develop a ranking system on a scale of 1-10 is
somewhat doubtful. Overlays can be used to predict spatial impacts, and the Sorensen
network is useful in tracing through indirect impacts.
Choice of prediction methods
The nature and choice of prediction methods do vary according to the impacts under
consideration, and Rodriguez-Bachiller with Glasson (2003) have identified the following
types:
• Hard-modelled impacts: areas of impact prediction where mathematical simulation
models play a central role. These include, for example, air and noise impacts. Air
pollution impact prediction has been dominated by approaches based on the so-called
“Gaussian dispersion model” which simulates the shape of the pollution plume from
the development under concern (Elsom 2001).
• Soft-modelled impacts: areas of impact prediction where the use of mathematical
simulation modelling is virtually non-existent. Examples here include terrestrial
ecology and landscape. Terrestrial ecology depends very much on field sample survey
for plant and animal species, where the expert's perception of what requires sampling
plays an important role (Morris & Thurling 2001). Perception is also important in
landscape assessment, but simple photomontages, and the use of GIS, can help in the
prediction of impacts (Therivel 2001, Wood 2000). Figure 5.4 provides an outline of
key steps in landscape assessment.
• Mixed-modelled impacts: areas of impact prediction where simulation modelling is
complemented (and sometimes replaced) by more technically lower-level approaches.
Traffic impacts make considerable use of modelling, but often with some sample
survey input. Socioeconomic impacts may use simple flow diagrams, and
mathematical models (as in Figures 5.2 and 5.3) particularly for economic impacts, but
they tend to build a great deal on survey methods and expert judgement. This is
particularly so with regard to social impacts.
When choosing prediction methods, an assessor should be concerned about their
appropriateness for the task involved, in the context of the resources available (Lee
1987). Will the methods produce what is wanted (e.g. a range of impacts, for the
appropriate geographical area, over various stages), from the resources available
(including time, data, range of expertise)? In addition, the criteria of replicability (method
is free from analyst bias), consistency (method can be applied to different projects to
allow predictions to be compared) and adaptability should also be considered in the
choice of methods. In many cases, more than one method may be appropriate. For
instance, the range of methods available for predicting impacts on air quality is apparent
from the 165 closely typed pages on the subject by Rau & Wooten (1980). Table 5.2
provides an overview of some of the methods of predicting the initial emissions of
pollutants, which, with atmospheric interaction, may degrade air quality, which may then
have adverse effects, for example on humans.